A Comparison Between Men Charged with Domestic and Nondomestic

Anasseril E. Daniel, MD, MRC Psych, and William R. Holcomb, PhD

Two hundred thirteen males charged with mur4er and who received pretrial psychiatric evaluations were divided into two groups of domestic and nondomestic homicide defendants. Demographics, developmental and fam­ ily background, prior criminal records, victim characteristics, and psychi­ atric status at the time of the crime were used to compare these groups. Those who were charged with intrafamilial homicide tend to be older, have a more stable adjustment in the community, but have more evidence of early childhood behavioral problems. Also, they are more likely to have committed prior crimes against persons. Two psychosocial stressors which were found to be important were the recent loss of employment and recent release from a psychiatric hospital. Post-offense suicidal attempts and psychotic/bizarre behavior were more prevalent for the intrafamilial mur­ derers. Implications of these results for forensic psychiatry and future directions of research are discussed. The significance of violence in the family as a major public health issue in the U.S. was highlighted by the appointment of the President's Task Force on Family Violence (1983). Although violence between family mem­ bers is rarely viewed as criminal, intrafamilial homicide is a well-defined entity and has been investigated extensively by criminologists. However, the psychiatric literature on homicide within the family seems to be sparse. within the family constitute one-third of all in the U.S. Spouses kill spouses in 50 percent of intrafamilial murders; in the rest, parents kill children, children kill parents, and other relatives kill one another. Several types of family are recognized: (1) (murder of a child by parents); (2) l (murder of the newborn child); (3) which is either (murder of by a child) or (murder of father); (4) (spouse murder); (5) (sibling murder); and (6) familicide2 (murder of the entire family). l 3 A review of specific categories - of intrafamilial murder suggests that

Dr: Daniel is an asaociate profe880r of peychiatry, University of Mi880uri, School of Medicine, and chief of staff, Mid-Mi880uri Mental Health Center, Columbia, MO. Dr. Holcomb is clinical aaaistant Professor of psychiatry, University of Mi880uri, School of Medicine, and the coordinator of Treatment Services, Mid-Mi880uri Mental Health Center, Columbia, MO. This paper was presented at the 15th Annual meeting of the American Academy of Psychiatry and the Law in N8888U, Bahamas, 1984. Address reprint requests to Dr. Daniel.

Bun Am Ac.d Paychlatry lIIw, Vol. 13, No.3, 1985 233 Daniel and Holcomb most perpetrators have a serious mental illness. Resnick' demonstrated in his review of the literature on filicidal parents from 1751 to 1967 that 70 percent of the filicidal were psychiatrically ill, most with a diagnosis of schizophrenia. Daniel and Harris3 reported that 87 percent of the mothers who killed their children were psychotic. Among patricidal killers, the most common diagnosis seems to be paranoid schizophrenia.4 While an associa­ tion seems to exist between psychosis and specific categories of domestic homicide such as matricide, filicide, or patricide, the cause and effect of such a relationship remain unknown. In contrast to matricidal, filicidal, and patricidal offenders, murderers of newborn children and spouse murderers seem to be less severely psychiat­ rically disturbed. While systematic and controlled studies of spouse murder are lacking, most of the relevant findings are derived from case reports that address only the dynamic aspects. Case reports2 indicate that the most likely pathologic state among famil­ icidal offenders seems to be an altered state of consciousness. Some are depressed and suicidal and attempt to destroy the family unit totally with an altruistic motivation. Occasionally, is committed with cold and calculated premeditation. Finally, occasional case reports attempt to link family murder to organic disorders such as hypoglycemia,S somnam­ bulism,6 and abnormal EEG patterns. 7 The purpose of our study was to compare the characteristics of domestic versus nondomestic homicide offenders with reference to demographic data, criminal profile, victim characteristics, family problems, and psychiatric diagnoses. We hypothesize that male domestic murderers have distinctive features when contrasted with nondomestic murderers.

Method Two hundred thirteen males charged with murder who received pretrial psychiatric evaluations were examined to study domestic and nondomestic homicide. The sample consisted of all homicide defendants admitted con­ secutively to a midwestern maximum security forensic unit from 1976 to 1981. It is to be noted that the facility admits defendants from anywhere within the state, including rural and metropolitan areas. Eighty percent of all evaluations of homicide defendants performed by the State Department of Mental Health are done at this facility. According to the Federal Bureau of Investigation Uniform Crime reports, this sample represents approxi­ mately 15 percent of all murder and non-negligent committed in the state over this six-year study period. Data for the study included interview notes, reports to the court, psycho­ logical reports, police reports, detailed social history, and, frequently, au­ topsy reports of the victims. We classified the sample into a domestic (44)

234 BuD Am Aced Plychilitry Law, Vol. 13, No.3, 1985 Domestic and Nondomestic Homicide and a non domestic group (169). The domestic defendant was defined as anyone who had a primary relationship with the victim and was living with the victim for more than six months. This group consisted of those who were charged with killing of wives (18), fathers (3), mothers (3), grandparents (2), children (5), live-in homosexual partner or girl friend (7), and other family members (4). The nondomestic group consisted of 169 defendants who were charged with killing of strangers, acquaintances, and victims of another crime such as robbery and . Statistical comparisons were made using analysis of variance and chi­ square tests of difference.

Results General Characteristics The sample of 213 defendants ranged in age from 15 to 74 years. While the mean age of the domestic homicide group was 35.04 ± 12.07, the mean age of the nondomestic group was 26.89 ± 10.2 (F = 18.79, P < .000 1). However, there were no significant differences between the two groups in regard to race, ratings of job skills, educational level, and the number of prior arrests and convictions. Analysis of differ­ ences on degree of urbanization of the defendant's county of residence based on 1980 census ratings indicated no differences between the two groups. Table 1 presents select demographic characteristics of the sample. Those who had committed either personal or mixed personal and property crimes in the past compared with only property crimes were more often found in

Table I. Comparison of Select DeJllOlP'1lpbic Characteristics of Domestic Homicide Offenders with NODdomestic Homidcle Offenders Domestic Nonciomestic Slpifkance N 'Yo N 'Yo x' df P Race White 34 77.27 110 65.09 2.36 2 NS Nonwhite 10 22.73 59 34.91 Marital status Divorced 5 11.90 29 17.47 Single 13 30.95 92 55.42 Widowed 13 30.95 2 1.20 45.85 4 0.0001 Separated 3 7.14 14 8.43 Married 8 19.05 29 17.47 Type of prior crimes Property 6 27.27 69 53.49 Person 9 40.91 12 9.30 16.25 2 0.0003 Mixed 7 31.82 48 37.21 Military history None 23 53.49 114 72.15 5.42 2 0.06 Bad discharge 5 11.63 11 6.97 Good discharge 15 34.88 33 20.89 Unemployed 24 55.81 105 66.04 1.5 NS

Bun Am AC8d Psychiatry l.8w, Vol. 13, No.3, 1885 231 -- Daniel and Holcomb the domestic groups. In regard to marital status, more of the nondomestic defendants were single at the time of the crime. Fifty-three percent among the domestic group and 72 percent of the nondomestic group did not have a military history. This result may partially be explained by the age differ­ ence between groups, with the older domestic group being more likely to have had a military record. Among those who had been in the military, the domestic group more often tended to have a good discharge. Finally, there were no differences between the groups on whether the offender was unemployed at the time of the murder. Fifty-six percent of the domestic group were unemployed at the time of the murder and the nondomestic group had 66 percent unemployment. Criminal Profile Eighty percent of those involving a family member occurred in either the victim's and/or offender's home compared with only

Table 2. Comparison or Data 011 Crime, Victim, Motive, aDd Post~fJense Behavior Between Domestic aDd Nondomestlc Homicide OfI'enden Domestic Nondomestlc N % N % Type of murder charged Capital 13 29.55 92 54.44 First degree 2 4.55 43 25.44 Second degree 25 56.82 33 19.53 40.17 3 0.0001 Manslaughter 4 9.09 1 0.59 Weapon/method of killing Rifle 3 6.82 17 10.06 Pistol 12 27.27 42 24.85 Shotgun 9 20.45 31 18.34 0.729 5 0.91 Other weapons I 2.27 6 3.55 Stabbing 10 22.73 38 22.49 No weapon 9 20.45 35 20.71 Motive Preplan ned 3 7.14 12 7.32 Self-defense 8 19.05 25 15.24 Accidental 5 11.90 13 7.93 Sex related 1 2.38 19 11.59 18.27 6 0.005 During a crime o 0.00 36 21.95 During arguments 11 26.19 21 12.80 No motive 14 33.33 38 23.17 Place of murder Victim/offender's home 35 83.33 56 34.15 Place of business 1 2.38 32 19.51 Street o 0.00 28 17.Q7 36.08 5 0.0001 Car 4 9.52 16 9.76 Other 2 4.76 32 19.21 Sex of victim Female 31 70.45 49 30.29 Male 13 29.55 113 69.75 23.55 0.0001 Behavior following murder Attempted 7 16.67 17 10.40 Psychotic 16 38.10 24 14.82 15.09 3 0.0001 Normal 19 45.24 121 74.69

Bun Am Aced Paychietry Lew, Vol. 13, No.3, 1915 Dome.tic and Nondome.tic Homicide 34 percent of the nonfamilial group. Besides home of the offenders or victims, the most frequent location of the nonfamilial killings were places of business, followed by other locations such as in the street or in a car. Weapons Table 2 shows that there were no significant differences be­ tween the two defendant groups in type of weapon used. However, it is interesting to note that 24 of 44 (55.5 percent) of the domestic group and 90 of 169 (53.24 percent) used a fire arm. Victim Profile Twenty-nine (66 percent) of the victims of domestic killers were adults (22 females and 7 males) and 34 percent were children (9 girls and 6 boys). Seventy-three percent of the nondomestic group were adults (35 females and 83 males) and 27 percent were children (14 girls and 30 boys). Analysis shows that victims of intrafamilial killers were predom­ inantly females (70 percent) in contrast to nonfamilial murder victims who were mostly males (70 percent). This suggests a significant risk for females in family violence and for males in nonfamily violence. Further analysis of age by sex revealed that adult females constituted 50 percent of all victims in the domestic group but only 27 percent of the nondomestic group (x2 = 10.93, df = I, p < .001). Although approximately one-third of all victims were children, they were not found at any greater risk in either group. Study of racial composition of victims showed that 25 percent of intrafamilial victims were black and 75 percent were white, which was not significantly different from nondomestic homicides. Multiple victims were involved in only two of the domestic homicides while there were multiple victims in fifteen of the nondomestic homicides. Developmental History Complete data were not available in regard to family and early childhood variables. However, 21 (48 percent) domestic killers and 61 (36 percent) nondomestic killers had a disrupted family during early childhood due to absence of a parent (see Table 3). Seven from the domestic group and 39 from the nondomestic group were physically abused, but this difference did not reach statistical significance. Interestingly, early childhood behavioral problems such as temper tantrums (p < .08), fights in school (p < .05), and juvenile delinquency (p < .03) were more often found among the domestic group. Psychosocial Stressors Several environmental stressors were specifically studied since different psychosocial stressors may play an important role in the precipitation of homicidal aggression. Domestic offenders were more likely to have experienced a recent loss of a job (p < .01) and a recent past hospitalization (p < .04). However, there were no differences between the two groups in regard to recent deaths in the family, recent release from Prison, financial loss, and marital separation. Post-offense Behavior Our data suggest that while 17 percent of killers of family members attempted to kill themselves after the homicide, this lull Am AC8d Psychi8try Law, Vol. 13, No.3, 1885 237 Daniel and Holcomb

Table 3. Comparison of Selective Developmental Data Domestic Nondomestic Significance 2 Item N % N % x df P Disrupted family 21 58.33 61 40.13 3.92 .047 Abused as a child 7 22.58 39 37.68 2.46 0.11 (NS) Father absent 18 50.00 64 45.71 0.21 NS Mother absent 4 12.12 35 25.18 2.59 0.10 (NS) Temper tantrums 41 93.18 140 82.89 2.95 0.08 Fights in school 39 88.64 127 75.15 3.69 0.05 Juvenile delinquency 34 77.27 102 60.36 4.32 0.03

behavior was present only among 10 percent of the nondomestic offenders. Furthermore, psychotic and bizarre behaviors immediately following the murder, as reported by the police records, tended to be higher among domestic killers. Twenty-five percent of the nondomestic killers exhibited abnormal behavior immediately after the homicide compared with 55 percent of the domestic group, suggesting that the domestic group may have been significantly more disturbed. Use of Drugs and Alcohol at the Time of Offense Although there were no significant differences between the two groups regarding alcohol or drug abuse at the time of the crime, use of any combination of intoxicants was slightly higher among the nondomestic killers. Thirty-two percent of the domestic group and 30 percent of the nondomestic group were intoxicated with alcohol at the time of the crime, whereas 10 percent of the domestic group and 17 percent of the non domestic group were under the influence of drugs other than alcohol. In addition, 15 percent of the domestic group and 20 percent of the nondomestic group were intoxicated with a combi­ nation of drugs and alcohol. Psychiatric Diagnosis A greater number of the domestic killers received psychotic diagnoses (34.09 percent) compared with nondomestic killers (11.67 percent) as a result of pretrial evaluations. Forty-six percent of the nondomestic group received alcohol and/or drug abuse diagnoses compared with 25 percent among the domestic group. Seven percent of the domestic group had anxiety (neurotic) disorders compared with 2 percent in the domestic group. Twenty percent of the domestic group were diagnosed as having antisocial personality disorder contrasted to 16 percent of non­ domestic defendants.

Discussion The generalizability of the findings reported in this article is limited by the highly selective nature of the population under study. The domestic homicide offenders consisted of 21 percent of all accused offenders referred for psychiatric assessment. Based on national statistics, one would expect a higher frequency of family murders in the state during the study period.

238 Bull Am Acad Psychiatry Law, Vol. 13, No.3, 1985 Domestic and Nondomestic Homicide One plausible explanation for this discrepancy may be that not all homicide offenders had psychiatric evaluations or had them at other facilities. Investigators have classified homicide into several types according to their theoretical perspectives. Perpetrators' psychopathology,S personality type,9 motivational and dynamic factors,1O victim-offender relationship, criminal and epidemiological characteristics, II and legal considerations have all formed the basis for such classifications. Categorization of homicide into familial and nonfamilial type may further aid in our understanding of family violence and consequently lead to systematic predictive and preven­ tive research of families at risk. While most of the violent crimes are committed by males between 15 and 24 years 01d,12 those who commit family violence tend to be older. Evidence seems to indicate that domestic homicide offenders tend to be more stable and able to adjust in the community over a long period of time. They are more likely to have a good discharge from the military when compared with nondomestic homicide offenders who served in the military. Also, they were slightly more likely to be employed at the time of the murder. There is less evidence of early family disruption caused by absence of one or both parents. On the other hand, there is more evidence in the early developmental history of childhood behavioral problems such as temper tantrums, fights in school, and juvenile delinquency. In addition, they were more likely to have committed prior crimes against persons than the nonfamilial murderers. The picture which emerges is of a person who has a propensity toward violence which episodically erupts throughout the history but who nevertheless is able to adjust marginally and establish relationships with others, obtain and hold jobs, and successfully serve in the military. The findings of higher frequency of psychotic diagnosis as well as bizarre behaviors and suicide attempts during the immediate post-offense period suggest that domestic offenders as a group were significantly more disturbed. l These findings are consistent with the previous studies ,3,4 on specific categories of intrafamilial homicide. This greater degree of psychopathology could have been precipitated by recent psychosocial stressors. Loss of employment and release from psychiatric hospitals tended to precede intra­ familial homicide. Previous studies13 reported that suicide attempts among murderers would be higher than among the general population. While the frequency of suicide attempts in our study among the nondomestic offenders generally correspond with that of the psychiatric population, it is signifi­ cantly higher among the intrafamilial offenders. Suicide attempt after murder often reflects intense guilt feelings of the offender. Perhaps the ambivalent love-hate relationship l4 between the offender and the victim is likely to be intensified in the close confines of the family.

Bull Am AC8d Psychiatry Law, Vol. 13, No.3, 1985 Daniel and Holcomb The homicidal event itself deserves further consideration. Both capital and first-degree murder require premeditation, planning, or an intent to kill during the commission of a felony. It is interesting to note that defendants of domestic homicide were significantly more often charged with second-degree murder or manslaughter. This difference may reflect either a bias of the criminal justice system against charging killers of family members with first-degree or capital murder, or an actual lack of preplan­ ning or premeditation in such homicides. Analysis of motives for killing lend support to the latter point of view. These findings further suggest that the lack of clear motive, criminal intent or planning, and killing in self­ defense and during arguments seem to be the hallmarks of intrafamilial homicide. Another significant finding of this study is that adult females seem to be at higher risk to be victims of violence within the family when a male perpetrator is involved. Although precise information regarding the quality of marital relationship is unavailable, it is likely that marital discord and spouse abuse are common. Contrary to our expectation, child victims were more or less evenly distributed between the two groups. Contrary to previous studies, there were no racial differences between the two groups. Nor were there any difference in the types of weapons used, with more than one-half of all murders resulting from the use of guns. Thus, overall findings of our study support the view that domestic homicide offenders, homicide events, and victims are quite different from those of nondomestic homicides. Even though this population of men charged with murder was divided into domestic and nondomestic offenders, a review of these cases clearly shows the heterogeneity of the group of domestic homicide offenders. Future research should attempt to develop a typology of violent family offenders by utilizing reliable psychological measurements, detailed social history, level of social functioning informa­ tion, recent psychosocial stressors, and medical and psychiatric status. Such a typology based upon an analysis of the interaction of individual and environmental variables would contribute substantially to our understand­ ing of family violence beyond the variables we have identified in this study.

1. Resnick PJ: Murder of the new born: A psychiatric review of neonaticide. Am J PlychiaUy 126:141"-1420,1970 2. Malmquiat CP: Paychiatric aapecta of familicide. Bull Am Acad Paychiatry Law 8:298-304, 1980 3. Daniel AE, Harria PW: Female homicide offendera referred for pretrial plychiatric examination: A deecriptive atudy. Bull Am Aced Plychiatry Law 10:261-269, 1982 4. Sadoff RL: Clinical obeervationa of parricide. PBychiatr Q 45:65-69, 1971 5. Hill D, Sargant WA: A cue of matricide. Lancet 1:526-527, 1943 6. Podollki E: Somnambulistic homicide. Am J Plychiatry 212:191-192,1964 7. Winfield DL, Ozturk 0: Electroencephalographic findinp in matricide. Oil Nerv 8yat 20:176- 178, 1959 8. Tanay E: Psychiatric Itudy of homicide. Am J Psychiatry 125:1252-1258, 1969

240 lui Am Acad Psychiatry Law, Vol. 13, No.3, 1185 Dome.tic and Nondom.stic Homicide

9. Cole KE, Fisher G, Cole SS: Women who kill: A sociopsychologicalstudy. Arch Gen Psychiatry 19:1-8, 1968 10. Revitch E, Schlesinger LB: Murder: Evaluation, classification and prediction, in Violence­ Perspectives on murder and aggression. Edited by Kutash IL, Kutash SB, and Schlesinger LB. San Francisco, Jossey-Bass Inc., 1978 11. Jason J, Strauss LT, Tyler CW: A comparison of primary and secondary homicides in the United Stetes. Am J EpidemioI117:309-31., 1983 12. APA Task Force Report 8: Clinical Aspects ofthe Violent Individual. Washington, ~C, American Psychiatric Association, 197. 13. Franke RH, Thomas EW, Queenen AJ: Suicide and homicide: Common sources and consistent relationships. Soc Psychiatry 12:a9-156, 1977 14. West OJ: Murder followed by Suicide. Cambridge, MA, Harvard University Press, 1966

lull Am Acad Psychilltry I.8w, Vol. 13, No.3, 1985 241